This proposal is designed to address the increase in drug use among adolescent athletes by studying a school-based version of the random, no-advance warning drug testing program used by the United States Olympic Committee (USOC). High school athletes are a large group, comprising 50 percent of their school's enrollment. They have a high rate of substance abuse behaviors similar to the general school population, and an even higher use of 'ergrogenic' (athletic enhancing) drugs. Recognizing the high rate of substance abuse among young athletes and their 'role model' effect on other students, the U.S. Supreme Court recently upheld an Oregon School Districts' policy to randomly drug test students engaged in school-sponsored sports. Drug testing has the potential to deter adolescent substance abuse. It is gender-neutral, without ethnic bias and provides a potentially powerful environmental influence. However, despite its legality and theorized effectiveness, schools are implementing drug surveillance without the benefit of randomized, prospective efficacy research. Focusing on adolescent athletes provides a unique opportunity to study the prevention effect of drug testing. All sports teams in 24 schools who agree to implement mandatory testing as school policy but have never implemented this policy, will be randomly assigned by school, to three years of either: 1) random, no-advance warning drug testing or 2) a 3-year control period without testing. Selection of students for drug testing will be random, with no exclusions for having been previously tested. State-of-the-art testing will include physician specimen collectors under the direction of research physicians (PI and Co-I) who are Certified USOC Drug Surveillance Crew Chiefs with specimen analysis at the UCLA Olympic Laboratory using the most accurate analytical techniques to minimize false negative (reducing policy integrity) and false positive (mislabling students) results. Confidential questionnaires will be completed by student-athletes twice yearly to assess risk and protective factors for drug use and asess self-reported substance abuse. The role model effect of the surveillance program on nonathletes' drug use will be assessed twice yearly by anonymous survey. We will determine the effect of drug testing policy on: 1) adolescent drug use mediators, 2) actual drug use behaviors of student-athletes and their non-athlete peers, and 3) potential gender and demographic differences. Reliability of subjective questionnaire responses will be assessed by comparisons with objective drug test results. Study findings will assist school districts and education agencies evaluate, guide, and implement future drug prevention policy decisions.